Quantifying implicit democracy

Abstract

The dissertation argues that standard measures of democracy (the Polity IV, ACLP databases) fail to take into account the potential ongoing influence of the previous authoritarian regime in the economic sphere of the country after an explicit democratic transition has taken place. It proposes a novel quantitative approach to test for the presence of such implicit influence.

The proposed method uses daily stock market data and exploits exogenous shocks (health shocks to the incumbent, commodity price shocks or assassinations) to each individual politico-economic system studied. In addition to financial data, the second major component to the analysis is the construction of a country specific connections dataset (each firm is identified as connected to the military, state, incumbent or the opposition) that relies on an extensive reading of historical and current secondary literature and news search.

The key hypothesis tested is that where implicit power has been transferred from the preceding authoritarian regime to elected officials, a connection to the former should not offer a relative safe haven from stock market turmoil in times of uncertainty over government turnover. If significant evidence is found for insulation extended by the former, subject to ruling out alternative explanations, it is argued that there is a shortfall in the transfer of implicit power. The dissertation takes a first step in identifying and empirically studying three different developing politico-economic settings, namely Turkey, Thailand and Pakistan. The empirical results show that while there was a robust shortfall in the transfer of implicit power in Turkey, the Thai polity was successful in neutralizing the potential implicit effects of the surviving elements of the preceding authoritarian regime in the early part of this decade. The finding is shown to be robust to alternate coding, alternate mechanisms, industry fixed effects and various important controls for firm characteristics. In contrast, the empirical analysis fails to find robust evidence for the presence/absence of implicit power of the authoritarian regime for the case of Pakistan. The divergent result from the qualitative literature on the topic is attributed to the scarce and non-transparent nature of the stock market data in Pakistan.